Snake oil and bulldust.
Serge
Amen brother!
Even reading 20+ year old papers on the subject doesn't prove anything or change my mind on the matter. Anybody who's wanting to make a quick buck on snake oil is willing to say anything to make people believe in their product(s). Not I...
"This is the sort of thing which your tax dollars are funding."
Tax dollars support our Mickey Mouse research institute:
Down Is Up - Weird Science in Action - Photo Gallery - LIFE
Tax dollars support our Mickey Mouse research institute:
Down Is Up - Weird Science in Action - Photo Gallery - LIFE
I didn't think tubes were made of soft glass anymore (50+ years). Thought new tubes (50 years old or less) were made with borosilicate glass . Makes this whole porous soft glass leak issue a moot point.
Espcially since low rpm cooling fans have been used on tube gear which is rack mounted for decades.
I still think spending 4 bucks on a metal low rpm cooling fan to cool ten tubes is a better value than spending 170 bucks to cool 10 tubes. forced air cooling is always better. no need to argue the point. An air cooled motorcycle will over heat if left standing still, but will cool very well once the bike moves. Forced air cooling.
I do like using a metal collar to reduce microphonics in my V1 tubes. I have a few dozen of those aluminum collars that have the 4 fingers that grip the tube. It transfers what little heat the preamp tubes produce while reducing microphonics. I bout a few dozen on ebay for 10 bucks.
Espcially since low rpm cooling fans have been used on tube gear which is rack mounted for decades.
I still think spending 4 bucks on a metal low rpm cooling fan to cool ten tubes is a better value than spending 170 bucks to cool 10 tubes. forced air cooling is always better. no need to argue the point. An air cooled motorcycle will over heat if left standing still, but will cool very well once the bike moves. Forced air cooling.
I do like using a metal collar to reduce microphonics in my V1 tubes. I have a few dozen of those aluminum collars that have the 4 fingers that grip the tube. It transfers what little heat the preamp tubes produce while reducing microphonics. I bout a few dozen on ebay for 10 bucks.
Last edited:
About 30 bucks in hand tools, probably half already own: You can have
as many chainmaille coolers as you got time to spare. 6BQ5/EL84/6P14P
take about 2hrs, maybe less, especially if you get organized about it.
You need:
Two sets of pliers (Home Despot or Lowes Life) covered with gorilla glue
(the polyurethane glue is only if you care about not leaving tool marks).
One set of vice grips and a Wall Mart knitting needle for coiling mandrel.
Turn the vice grips and form coil by hand, do NOT use a power drill!
One set of tungsten carbide tile nippers, and dremel tool to modify the
stops (that part of the tool is not tungsten) to close down enough to
score 18 ga for clean breaking.
Breaking rings off a coil with the nippers is... I'd probably have to snap
you a picture, its not so easy to describe the technique in words. You
can use any cutting tool, but to avoid deforming the rings is the main
reason you will want to score and break them with nippers.
The toughest problem with aluminum is that its too damn soft, and will
easily deform in unintended ways.
If you got more money than time to throw at it, there's places online to
buy black anodized aluminum rings... But making the rings isn't really all
that time consuming. 15min out of your 2hrs per heatsink.
The purpose of DIY is not to buy stuff that you have the skill to make.
as many chainmaille coolers as you got time to spare. 6BQ5/EL84/6P14P
take about 2hrs, maybe less, especially if you get organized about it.
You need:
Two sets of pliers (Home Despot or Lowes Life) covered with gorilla glue
(the polyurethane glue is only if you care about not leaving tool marks).
One set of vice grips and a Wall Mart knitting needle for coiling mandrel.
Turn the vice grips and form coil by hand, do NOT use a power drill!
One set of tungsten carbide tile nippers, and dremel tool to modify the
stops (that part of the tool is not tungsten) to close down enough to
score 18 ga for clean breaking.
Breaking rings off a coil with the nippers is... I'd probably have to snap
you a picture, its not so easy to describe the technique in words. You
can use any cutting tool, but to avoid deforming the rings is the main
reason you will want to score and break them with nippers.
The toughest problem with aluminum is that its too damn soft, and will
easily deform in unintended ways.
If you got more money than time to throw at it, there's places online to
buy black anodized aluminum rings... But making the rings isn't really all
that time consuming. 15min out of your 2hrs per heatsink.
The purpose of DIY is not to buy stuff that you have the skill to make.
Last edited:
Tube-coolers.....
Load of old cods!
Only transmitting tubes and that ilk made with metal anodes extending outside the glass need cooling--They are Designed for that purpose...
But all the usual suspects used in home audio--The ONLY reason you would want to place a can of any sort on 'em is for Screening!
An amp designed properly, allowing sufficient room between high diss valves will have no trouble.
Tube coolers.......Utter Cods!--Only thing they are good for is making the vendor money!
Load of old cods!
Only transmitting tubes and that ilk made with metal anodes extending outside the glass need cooling--They are Designed for that purpose...
But all the usual suspects used in home audio--The ONLY reason you would want to place a can of any sort on 'em is for Screening!
An amp designed properly, allowing sufficient room between high diss valves will have no trouble.
Tube coolers.......Utter Cods!--Only thing they are good for is making the vendor money!
Exactly!
Another fine point that the 20+ year old paper tends to leave out. ONLY mentioning the "soft glass" issue without letting people know that it is a non-issue these days make those people think that they need these silly useless coolers that much more.
It's all total snake oil with the sole purpose to make loads of money off the uneducated buyers. Typical this day and age, especially in the hifi world.
While I was going to contribute to this 'thread' it is clear that is so often the case the self-informed, opinionated 'experts' will seek to have their day, preferring to level specious accusation instead of actually learning something/anything.
Dealing with the comment two posts up: as they always were, conventional receiving tubes, of which the 'audio' types can perhaps be thought of as a sub-class, have historically been made with 'soft' glass as opposed to 'hard.' The latter being Pyrex™.
Perhaps the 'experts' here will find it amusing to learn that the Svetlana plant in St. Petersburg, Russia manufacture their KT88, EL34, 6550C and 6L6 with hard glass, i.e. PYREX™, envelopes perhaps the only maker in history to build 'consumer grade' parts to such a standard.
I was involved back in the days when Svet' was developing their 6550 and Eric Barbour, Geo, Badger and I talked about the longevity benefits of using hard glass.
Those who will undoubtedly doubt that Pyrex™ is in use there need only catch, say, an ordinary Pyrex™ measuring cup in the right light to see the greenish cast 'within' such glass and then catch any of the tubes just mentioned in the same light and observe the same cast.
Being the one who sorted David Manley's problems with the early KT90s, an issue with screen grid/anode geometry critical in beam tetrodes, I have lately communicated with Ei in respect of putting that excellent part in a hard envelope and fitting a rather better base. Cost and availability are for them apparently an issue.
While not having to do with tube coolers per se, it might be interesting to note that when working with Svet' on the 6550A/B/C I suggested the placement of a SAES continuous acting, strip getter directly on the hot anode where it would ongoingly scavenge any gases that might arise, contributing thereby to extended tube life
SAES is the little company who make such materials, along with materials suitable for the evacuation of amateur, hobby projects such as the colliders at CERN.
And speaking of gettering, when I had Svet' mod a metal ceramic triode then used as a voltage reg' pass device into what became their 3CX300A1, and my RB300-3CX a 300W triode, I insisted they double the number of what are called 'pill getters' within the envelope -- again, continuously acting getters but small, about the size of an Aspirin™.
Further, I had them replace the steel, outer anode ring with a solid copper part and then developed the low-velocity, air cooling fan/chimney arrangement you can see here:
PEARL Misc. D'loads
Or directly download from here:
http://www.pearl-hifi.com/06_Lit_Archive/07_Misc_Downloads/RB300_3CX_Data_Sheet.pdf
Earlier in this thread someone mentioned the use of a Russian 6S19P, a low mu (2.5-ish), high perveance single triode with a cathode about twice the size of that in a 12B4.
Parenthetically; I recently developed a series pass reg' with that part, a 6AU6 error amp with a current source as plate load and current sink based DC level shifting to get some 90% of the reg's output error voltages directly coupled to the error amp's control grid. 1N821s are used in series as the Vref. The screen has a trim arrangement to trim the reg's output to an exact value.
While working on that circuit I set the 6S19P to dissipate about 15W and measured its bare-bulb temperature, the bulb temp with a PEARL cooler and, finally with a fan and chimney.
Bare, the part's hot spot ran about 150ºC; with a cooler it was down to 100-110ºC and with the fan/chimney it dropped to around 60ºC. From a strictly thermal standpoint this will increase the SOA.
Finally, the 6S19P uses a continuously acting getter rather a flashed-one-time sacrificial getter . . . Highly unusual in a receiving tube and clever to boot.
Dealing with the comment two posts up: as they always were, conventional receiving tubes, of which the 'audio' types can perhaps be thought of as a sub-class, have historically been made with 'soft' glass as opposed to 'hard.' The latter being Pyrex™.
Perhaps the 'experts' here will find it amusing to learn that the Svetlana plant in St. Petersburg, Russia manufacture their KT88, EL34, 6550C and 6L6 with hard glass, i.e. PYREX™, envelopes perhaps the only maker in history to build 'consumer grade' parts to such a standard.
I was involved back in the days when Svet' was developing their 6550 and Eric Barbour, Geo, Badger and I talked about the longevity benefits of using hard glass.
Those who will undoubtedly doubt that Pyrex™ is in use there need only catch, say, an ordinary Pyrex™ measuring cup in the right light to see the greenish cast 'within' such glass and then catch any of the tubes just mentioned in the same light and observe the same cast.
Being the one who sorted David Manley's problems with the early KT90s, an issue with screen grid/anode geometry critical in beam tetrodes, I have lately communicated with Ei in respect of putting that excellent part in a hard envelope and fitting a rather better base. Cost and availability are for them apparently an issue.
While not having to do with tube coolers per se, it might be interesting to note that when working with Svet' on the 6550A/B/C I suggested the placement of a SAES continuous acting, strip getter directly on the hot anode where it would ongoingly scavenge any gases that might arise, contributing thereby to extended tube life
SAES is the little company who make such materials, along with materials suitable for the evacuation of amateur, hobby projects such as the colliders at CERN.
And speaking of gettering, when I had Svet' mod a metal ceramic triode then used as a voltage reg' pass device into what became their 3CX300A1, and my RB300-3CX a 300W triode, I insisted they double the number of what are called 'pill getters' within the envelope -- again, continuously acting getters but small, about the size of an Aspirin™.
Further, I had them replace the steel, outer anode ring with a solid copper part and then developed the low-velocity, air cooling fan/chimney arrangement you can see here:
PEARL Misc. D'loads
Or directly download from here:
http://www.pearl-hifi.com/06_Lit_Archive/07_Misc_Downloads/RB300_3CX_Data_Sheet.pdf
Earlier in this thread someone mentioned the use of a Russian 6S19P, a low mu (2.5-ish), high perveance single triode with a cathode about twice the size of that in a 12B4.
Parenthetically; I recently developed a series pass reg' with that part, a 6AU6 error amp with a current source as plate load and current sink based DC level shifting to get some 90% of the reg's output error voltages directly coupled to the error amp's control grid. 1N821s are used in series as the Vref. The screen has a trim arrangement to trim the reg's output to an exact value.
While working on that circuit I set the 6S19P to dissipate about 15W and measured its bare-bulb temperature, the bulb temp with a PEARL cooler and, finally with a fan and chimney.
Bare, the part's hot spot ran about 150ºC; with a cooler it was down to 100-110ºC and with the fan/chimney it dropped to around 60ºC. From a strictly thermal standpoint this will increase the SOA.
Finally, the 6S19P uses a continuously acting getter rather a flashed-one-time sacrificial getter . . . Highly unusual in a receiving tube and clever to boot.
Last edited:
Hmmm...
Actually, I've opened this thread in order to find heatsinks for 6С19П tubes. My conclusion is, I will copy that military heatsinks I have...
Thanks anyway!
Small-error correction:
I wrote:
"While working on that circuit I set the 6S19P to dissipate about 15W and measured its bare-bulb temperature, the bulb temp with a PEARL cooler and, finally with a fan and chimney."
when I should have said:
"While working on that circuit I set the 6S19P to dissipate about 15W and measured its bare-bulb temperature, the bulb temp with a PEARL cooler and, finally with the cooler augmented by a fan/chimney arrangement where I ran a 40mm, 12V Taiwanese-made, Sunon fan at dead quiet 6V. "
I wrote:
"While working on that circuit I set the 6S19P to dissipate about 15W and measured its bare-bulb temperature, the bulb temp with a PEARL cooler and, finally with a fan and chimney."
when I should have said:
"While working on that circuit I set the 6S19P to dissipate about 15W and measured its bare-bulb temperature, the bulb temp with a PEARL cooler and, finally with the cooler augmented by a fan/chimney arrangement where I ran a 40mm, 12V Taiwanese-made, Sunon fan at dead quiet 6V. "
Do you have expert opinion wether maille might be better than nothing or worse?
I'm aware its not likely to compare well against a professionally engineered sink.
However, it could be tailored to fit oddball tube shapes, coke bottles etc...
Not as difficult as that might sound, just time consuming...
-----
I have an 92mm Delta running at 15V, and can't hear any hum, hiss, click, pop
or music over the whine. Perfect! (in my PC, not my tube amp)
I'm aware its not likely to compare well against a professionally engineered sink.
However, it could be tailored to fit oddball tube shapes, coke bottles etc...
Not as difficult as that might sound, just time consuming...
-----
I have an 92mm Delta running at 15V, and can't hear any hum, hiss, click, pop
or music over the whine. Perfect! (in my PC, not my tube amp)
Attachments
Last edited:
tube coolers
My mono block amps, built in 1963, have 8ea el34 packed in a 6in x 6in area. Lots going on in a small area, but low speed super quiet computer fans have now replaced the original fans. I have no problems with heat, and the Svetlanas are still running strong after many years. This approach also extracts heat from the chassis to relieve heating problems within. Those fans are also variable speed so you can dial in only the amount of heat extraction needed and still keep the noise at a minimum. I believe the brand of fan is Compuman. Still working after 9 years, cheap too.
My mono block amps, built in 1963, have 8ea el34 packed in a 6in x 6in area. Lots going on in a small area, but low speed super quiet computer fans have now replaced the original fans. I have no problems with heat, and the Svetlanas are still running strong after many years. This approach also extracts heat from the chassis to relieve heating problems within. Those fans are also variable speed so you can dial in only the amount of heat extraction needed and still keep the noise at a minimum. I believe the brand of fan is Compuman. Still working after 9 years, cheap too.
Ken Peter wrote:
"Do you have expert opinion whether maille might be better than nothing or worse?
I'm aware its not likely to compare well against a professionally engineered sink.
However, it could be tailored to fit oddball tube shapes, coke bottles etc...
Not as difficult as that might sound, just time consuming..."
If your question is directed me, I don't have any 'expert' opinions I just do the work ;-)
I think your idea above is very clever. It will rid you of hot spots and the enturbulation of air flow around the maille is a good thing. A hint, paint your parts with some hi-temp, flat black paint, it'll make all the difference.
He also observed:
"Sadly, the quietest fans are sleeve bearing, and also the shortest lived..."
That is not so, read the very end of my paper on the RB300 and note that the supremely silent fans ETRI make are ball bearing and available off the shelf that way. And that little Sunon fan out of Taiwan is also a ball bearing part
"Do you have expert opinion whether maille might be better than nothing or worse?
I'm aware its not likely to compare well against a professionally engineered sink.
However, it could be tailored to fit oddball tube shapes, coke bottles etc...
Not as difficult as that might sound, just time consuming..."
If your question is directed me, I don't have any 'expert' opinions I just do the work ;-)
I think your idea above is very clever. It will rid you of hot spots and the enturbulation of air flow around the maille is a good thing. A hint, paint your parts with some hi-temp, flat black paint, it'll make all the difference.
He also observed:
"Sadly, the quietest fans are sleeve bearing, and also the shortest lived..."
That is not so, read the very end of my paper on the RB300 and note that the supremely silent fans ETRI make are ball bearing and available off the shelf that way. And that little Sunon fan out of Taiwan is also a ball bearing part
Yep, and I'll bet those Svets' are hard glass.
I didn't mention that the large, glass transmitting tubes use hard glass well. I once measured the envelopes on the 813s in my heavily modified Altec 260s at 550ºF with the graphite plates -just- running dull cherry in the dark.
Whether or not the now ubiquitous Sino 2- and 8xx-series parts use hard glass I do not know but I suspect not. That said I'm happy to shown wrong . . .
There are magnetic floating bearing fans too. Sunon makes some, others too. Maybe more for reliability, but they should be dead quiet as far as the bearings go.
The RCA TT3 transmitting tube handbook says that soft glass softens at 625 degrees C and hard glass softens at 750 degrees C. The majority of Sweep tubes I've seen temp ratings on say 200 to 260 degrees C Max bulb temp. 220 C being the most common figure.
I just looked at the RCA 813 data sheet and it says in the notes at the end that nothing should touch the glass while in operation. And come to think of it, the only place I've seen the slip on MIL coolers was for regulator tubes in a tightly packed chassis.
The RCA TT3 transmitting tube handbook says that soft glass softens at 625 degrees C and hard glass softens at 750 degrees C. The majority of Sweep tubes I've seen temp ratings on say 200 to 260 degrees C Max bulb temp. 220 C being the most common figure.
I just looked at the RCA 813 data sheet and it says in the notes at the end that nothing should touch the glass while in operation. And come to think of it, the only place I've seen the slip on MIL coolers was for regulator tubes in a tightly packed chassis.